Effects of noise and mental task performance upon changes in cerebral blood flow parameters.

The objectives of this paper were to determine whether traffic noise influences the parameters of cerebral blood flow (CBF) measured by functional transcranial Doppler sonography (fTCD) during the performance of mental tasks, and to see whether impact of noise on CBF changes with age. The study comprised 36 healthy volunteers, 22 women and 14 men, aged 25-49 years. The fTCD was performed using a fixed 2-MHz probe, aiming for an evaluation of mean velocity (MFV) and the pulsatility index (PI) in the middle cerebral artery (MCA) on both sides. Subsequently, fTCD was monitored: At rest; during performance of the Paced Auditory Serial Addition Test (PASAT); during exposure to traffic noise; and during concomitant exposure to noise and PASAT performance. MFV and PI were compared for particular conditions and correlated with age. During exposure to noise, flow parameters did not change significantly. PASAT performance in silence increased MFV and decreased PI in MCA on both sides. During PASAT performance, on exposure to noise, MCV and PI changed significantly only in the left MCA. However, values of MFV were significantly lower during noise than in silence. Correlations with age were noted for velocities in the right MCA during PASAT performance in silence and for PI on both sides during PASAT performed in noise conditions. Noise impairs the CBF during mental tasks. A comparison of changes in CBF parameters correlated with age suggests that the involvement of the nondominant hemisphere in managing with noise effects increases with age.

Introduction

The issue of environmental noise and its influence on human health is gaining importance due to the intensification of traffic and increase in average population density. According to the World Health Organization (WHO), approximately 30% of the European Union population is exposed to traffic noise exceeding 55 dB.[1] Constant exposure to noise causes stimulation of the endocrine and autonomic nervous systems. Their response includes an increase in adrenaline and cortisol serum concentration and in dyselectrolytemia, which in turn cause an increase in vessel resistance, thus constituting a risk factor for cardiovascular diseases (i.e. hypertension, myocardial infarction).[2] However, knowledge of noise-induced systemic vascular reactions may not be directly applicable to cerebral vessels.

Other health consequences arising from exposure to noise include: sleep disturbances,[34] and noise annoyance, which shows one of its highest rates within the population in Poland according to SILENCE project data.[5] Fatigue and a lack of concentration, often associated with increased cortisol levels, are the most prevalent noise-related health problems.[67] Noise can impair cognitive performance, especially in those aspects bound to speech perception and its acoustic analysis. Such negative consequences are described especially in children, teenagers, and elderly persons.[8910]

Functional transcranial Doppler sonography (fTCD), a noninvasive method of great temporal resolution, can analyze parameters of cerebral blood flow (CBF) in the main intracranial vessels. Assuming that the diameter of the large cerebral arteries remains unchanged over time (even in the presence of various physiological stimuli), blood flow velocity (BFV) changes can be attributed to volume flow changes. Hence, the values of cerebral BFV may correspond with changes in the neuronal metabolism due to cerebral activation. So far, there have only been a few studies on healthy subjects, investigating objective changes in the resistance of cerebral vessels influenced by intensive auditory stimuli, including noise.

The aim of this study was, therefore, to estimate the influence of environmental noise upon parameters of CBF (assessed by means of fTCD) as well as upon cognitive performance in healthy volunteers with regard to demographic factors.

Methods

The study comprised 36 healthy, right-handed volunteers, 22 women and 14 men, aged 25-49 (mean: 33.7 years; standard deviation: 7.89). None of the subjects had any symptoms of neurological, psychiatric, or systemic disease, neither had they taken any medication on the day of examination or on the preceding day. Participants were recruited from among undergraduate students of the Medical University and the hospital staff (doctors and nurses). All the subjects provided informed consent prior to their inclusion in the study. The local Ethics Committee approved the protocol of the study.

First, fTCD was performed with the use of a Nicolet™ Sonara/tek™ Transcranial Doppler System (VIASYS Healthcare Inc., USA) with a fixed 2 MHz probe. The following parameters of the flow in the middle cerebral artery (MCA) were taken into account: Mean flow velocity (MFV), peak flow velocity (PFV), end diastolic velocity (EDV), systolic/diastolic velocity ratio (SD), the resistive index (RI) and the pulsatility index (PI). Following an initial rest phase in silence (2 min), fTCD was monitored on both MCAs during subsequent activation tasks separated by rest phases. The activation tasks were applied in the following order: Exposure to noise, performance of the mental task in silence, and performance of the mental task on exposure to noise. Each activation task lasted 3 min 15 s and the whole session took approximately 30 min.

The Paced Auditory Serial Addition Test (PASAT, 3-s version[11]) was used as the mental task. This test enables the assessment of the capacity and rate of information processing as well as sustained and divided attention. Because the test was performed four times overall during one session, versions A and B were used alternately in subsequent activation tasks for each studied side to minimize the learning effect. The sound level of PASAT signals was established at 65 dB sound pressure level (SPL) and increased accordingly when applied concomitantly with noise. The correctness of answers was marked during the task on the answer sheet. The final score equals the number of the correct responses out of the 60 serial addition results.

The traffic noise had been recorded in one of the main streets of the city (Wroclaw, Poland). The ½-inch measurement microphone with preamplifier (PCB model 378B02, PCB Piezotronics Inc., USA) was applied. The calibration was done with the acoustic calibrator model SV30 1000 Hz/94 dB (Svantek, Poland). The sound level varied in the range 45-85 dB SPL, and its frequency 50-5000 Hz (the authors’ own record and analysis) [Figure 1]. The noise was applied to the subjects using headset.

Figure 1

Frequency (Hz) and sound intensity level (dB) of traffic noise (the authors’ own record and analysis)

Blood flow parameters (FV, PI, RI) in both MCA were compared for particular conditions (activation tasks, rest phases) and age groups. The average values for each parameter were calculated for the last 30 s of the rest phases and for the middle 30 s of the activation task periods for each subject. The values of parameters obtained during activation periods were compared with those obtained during directly preceding rest phases [Figure 2].

Figure 2

(a) Changes of MFV, PI, and SD in the right MCA of a 25-year-old man (b) Changes of MFV, PI, and SD in left MCA in the same subject

To minimize the impact of varying insonation angle (which is not eliminated using the fTCD method) in the case of the MFVs, the relative increase from baseline to activation was calculated based on the following equation:

(MFVactivation - MFVrest)/MFVrest * 100

Statistical methods

The results were treated as significant if the P value was less than 0.05. In particular cases, a P value close to 0.05 was considered sufficient to observe trends. The normality of distribution was verified with the Shapiro-Wilk test. If a normal distribution and homogeneity of variances (checked with Levene's test) were stated, a comparison of the groups was performed with the parametric Student's t-test. If the parameter value distributions were significantly different from a normal distribution, the nonparametric Mann-Whitney U test was used for comparison. As a comparison of changes between parameters observed for the same subject but under different conditions, tests for dependent observations were used. In the case of normally distributed variables, the Student's t-test was applied. In other cases, differences were checked with a nonparametric sign test or the Wilcoxon matched pairs test.

For assessing dependencies between two variables, correlation coefficients were calculated and assessed. In the case of normally distributed variables a standard Pearson coefficient was used. In other cases, Spearman's rank correlation was used.

The statistical analysis was performed using STATISTICA v.8 statistical software (Statsoft Inc., USA).

RESULTS

The MCA could be identified and investigated by TCD method in all subjects.

TCD. Condition: Silence

In baseline conditions no significant differences in TCD parameters were noted between either sides or sexes. No correlation with age was noted for any analyzed parameter.

TCD. Condition: Noise

During exposure to noise, none of the analyzed parameters of blood flow changed significantly in comparison to the baseline (in the whole group of subjects). There was only a trend (P = 0.055) for an increase in heart rate (HR) during exposure to noise in the left MCA. No lateralization was noted. Age correlated negatively with HR change during noise and baseline conditions.

TCD. Condition: Silence and PASAT

All analyzed parameters of blood flow in the left MCA significantly changed during performance of PASAT in comparison with baseline values. All measured flow velocities (MFV, PFV, EDV) and the HR increased. SD decreased, as did PI and RI. The choice of statistical test used for the analysis was dependent on the fulfilment of the normality assumption. The MFV, PFV, and EDV parameters and HR were compared with Student's t-test for dependent samples. For PI and RI, the Wilcoxon matched pairs test was used [Table 1].

Table 1

Comparisons of flow parameters in right and left MCA during PASAT performance (P) and rest state (C2)

	Right side						Left side
	Mean	Std. Dv.	N	Diff.	Test statistic	P-level	Mean	Std. Dv.	N	Diff.	Test statistic	P-level
MFV C2	56.28	10.73					56.25	11.10
MFV P	59.58	10.85	36	−3.31	−2.76	0.01	60.71	12.11	36	−4.46	−3.85	0.00
PFV C2	93.27	16.98					92.20	15.16
PFV P	96.37	15.64	36	−3.09	−1.53	0.13	97.20	18.24	36	−5.01	−3.73	0.00
EDV C2	35.30	8.35					36.51	9.39
EDV P	38.22	9.21	36	−2.92	−3.94	0.00	39.78	9.56	36	−3.27	−3.33	0.00
PI C2	1.03	0.22					1.00	0.23
PI P	0.98	0.21	36	0.05	2.41	0.01	0.95	0.21	36	0.05	2.73	0.01*
RI C2	0.61	0.07					0.60	0.07
RI P	0.59	0.07	36	0.02	2.67	0.01	0.58	0.07	36	0.02	2.55	0.01*
SD C2	2.78	0.85					2.63	0.57
SD P	2.65	0.61	36	0.13	2.29	0.02*	2.51	0.49	36	0.11	2.55	0.01*
HR C2	81.60	12.67					80.55	10.07
HR P	90.02	13.71	36	−8.42	−5.91	0.00	89.34	14.80	36	−8.80	−5.17	0.00

MFV = Mean flow velocity, PFV = Peak flow velocity, EDV = End-diastolic velocity, PI = Pulsatility index, RI = Resistive index, HR = Heart rate. In cases where the Wilcoxon matched pairs test was used (with test statistic Z) a star (*) is placed next to the P value. In other cases, the Student’s t-test with test statistic t was used

The parameters analyzed for the right MCA in almost all cases showed significant changes during PASAT performance versus baseline. MFV and EDV as well as HR increased. PI and RI decreased during PASAT in silence. In the case of SD, which was analyzed using the nonparametric Wilcoxon matched pairs test, a significant decrease was found [Table 1]. No significant difference was noted in the case of PFV. The relative changes of MFV between both sides, calculated in order to avoid the influence of the insonation angle, were significantly higher on the left than on the right side (29%; P < 0.05). Significant correlations between age and the changes of flow parameters during PASAT performance were observed only for the right MCA (using the nonparametric Spearman correlation test) [Table 2], and they concerned only the velocities and not RI or PI.

Table 2

Correlations with age of the flow parameter changes in left and right MCA during PASAT test performance (P) in comparison to rest state (C2)

	Right side				Left side
	N	Spearman	t (N-2)	P-level	Spearman	t (N-2)	P-level
MFV P-C2	36	0.46	3.01	0.00	−0.14	−0.81	0.43
PFV P-C2	36	0.47	3.08	0.00	−0.15	−0.90	0.37
EDV P-C2	36	0.33	2.05	0.05	−0.03	−0.17	0.86
PI P-C2	36	0.16	0.92	0.37	0.10	0.61	0.55
RI P-C2	36	0.10	0.61	0.55	−0.06	−0.37	0.72
SD P–C2	36	0.17	1.03	0.31	0.08	0.49	0.63
HR P-C2	36	−0.27	−1.63	0.11	−0.22	−1.29	0.21

MFV = Mean flow velocity, PFV = Peak flow velocity, EDV = End-diastolic velocity, PI = Pulsatility index, RI = Resistive index, HR = Heart rate

TCD. Condition: Noise and PASAT

Most analyzed parameters of flow in the left MCA (except for PFV) changed significantly in comparison with baseline measurements: There was an increase in MFV and EDV, and a decrease in PI, RI, and SD ratios. No such changes were found for the right MCA. The HR increased significantly [Table 3]. The relative changes of MFV between both sides, calculated in order to avoid the influence of the insonation angle, were found to be significantly higher on the left than on the right side (41%; P < 0.05).

Table 3

Parameters of flow in the right and in the left MCA during PASAT performance on exposure to noise (PH) and rest state (C3)

	Right side						Left side
	Mean	Std. Dv.	N	Diff.	Test statistic	P-level	Mean	Std. Dv.	N	Diff.	Test statistic	P-level
MFV C3	56.21	9.74					55.81	10.99
MFV PH	57.90	11.91	36	−1.69	−1.41	0.17	58.10	11.73	36	−2.29	−2.60	0.01
PFV C3	92.80	14.98					92.34	14.80
PFV PH	93.09	16.06	36	−0.28	−0.21	0.83	93.85	16.39	36	−1.51	−1.14	0.26
EDV C3	35.66	8.51					35.80	9.27
EDV PH	35.11	10.82	36	0.55	0.55	0.59	37.51	9.39	36	−1.71	−2.51	0.02
PI C3	1.02	0.24					1.03	0.26
PI PH	1.00	0.23	36	0.01	0.54	0.59	0.97	0.21	36	0.05	2.50	0.01
RI C3	0.60	0.08					0.60	0.07
RI PH	0.61	0.09	36	−0.01	−0.51	0.61	0.59	0.07	36	0.01	2.36	0.02
SD C3	2.72	0.64					2.70	0.62
SD PH	3.53	3.58	36	−0.82	0.43	0.67*	2.61	0.57	36	0.09	2.22	0.03*
HR PH	81.51	12.03					81.00	12.05
HR C3	91.85	13.43	36	−10.34	−7.25	0.00	89.91	14.35	36	−8.91	−6.75	0.00

MFV = Mean flow velocity, PFV = Peak flow velocity, EDV = End-diastolic velocity, PI = Pulsatility index, RI = Resistive index, HR = Heart rate. In cases when the Wilcoxon matched pairs test (with test statistic Z) was used a star (*) is placed next to the P value. In other cases, the Student’s t-test with test statistic t was used

The increase in PI on both sides and RI on the left side correlated significantly with age [Table 4]. For the right side, there was a strong tendency for RI to increase with age (P = 0.07). Simultaneously, a correlation with age was found for the peak velocity changes, which rose significantly in the right MCA (P < 0.05), and a trend toward such an increase was observed for the left MCA (P = 0.07).

Table 4

Correlations with age of the relative flow parameter changes in left and right MCA during PASAT performance on exposure to noise (PH) in comparison to rest state (C3)

		Right side			Left side
	N	Spearman	t (N-2)	P-level	Spearman	t (N-2)	P-level
MFV PH-C3	36	0.11	0.64	0.53	0.12	0.68	0.50
PFV PH-C3	36	0.35	2.21	0.03	0.31	1.87	0.07
EDV PH-C3	36	−0.03	−0.19	0.85	−0.08	−0.46	0.65
PI PH-C3	36	0.38	2.37	0.02	0.35	2.16	0.04
RI PH-C3	36	0.31	1.89	0.07	0.34	2.09	0.04
SD P-C2	36	0.31	1.90	0.07	0.31	1.89	0.07
HR PH-C3	36	−0.02	−0.11	0.91	−0.38	−2.37	0.02

MFV = Mean flow velocity, PFV = Peak flow velocity, EDV = End-diastolic velocity, PI = Pulsatility index, RI = Resistive index, HR = Heart rate

TCD. Condition: PASAT in silence versus noise

The relative changes were calculated based on the following equation:

(MFVPASAT in noise - MFVPASAT in silence)/MFVPASAT in silence * 100

in order to minimalize the influence of the insonation angle. The relative values of MFV were significantly lower during noise than in silence (47.9%, P < 0.05 and 52.2%, P < 0.05 for the right and the left side, respectively).

PASAT. Condition: Silence versus noise

The mean PASAT result in silence (version A) was 52 points (out of 60) and on exposure to noise (version B) it was 53 points. The results obtained during the left MCA assessment, which was conducted in the second part of the whole testing session, were significantly better than those obtained during the right MCA evaluation (49.7 points vs 54.6 points for version A; 51.8 points vs 55.6 points for version B). The results of PASAT version B, which was applied during noise, improved in 25 subjects (70%), declined in 9 (25%), and remained unchanged in 2 (5%).

TCD in particular conditions. Sex and age differences

Males and females did not differ in their reactivity to noise: None of the analyzed parameters of blood flow changed significantly in comparison to the baseline for either gender. Significant differences were noted during PASAT performance. Males showed a significantly higher relative increase in MFV and EDV as well as a decrease in the PI and RI in the left MCA. During PASAT performance on exposure to noise, there was a significantly higher increase in MFV, PFV, and EDV on the left side in males than in females [Table 5]. No significant age-correlated changes were noted during exposure to noise. The changes of flow parameters during PASAT performance correlated significantly with age in a few cases: Velocities in the right MCA in silence [Table 2], and in PI on both sides, PFV on the right, and RI on the left in noise conditions [Table 4].

Table 5

Comparisons of relative to basal line changes of flow parameters between the sexes in right and left MCA during PASAT performance (P) and rest state (C2) and during PASAT performance on exposure to noise (PH)

	Right side				Left side
	Mean values		T-value	P-level	Mean values		T-value	P-level
	Females (%)	Males (%)			Females (%)	Males (%)
MFV P-C2	5	9	−1.10	0.28	5	14	−2.29	0.03
PFV P-C2	3	7	−1.06	0.30	4	8	−1.44	0.16
EDV P-C2	8	10	−0.63	0.54	6	18	−2.21	0.03
PI P-C2	−4	−4	0.05	0.96	0	−10	2.21	0.04
RI P-C2	−2	−3	0.34	0.74	0	−6	2.16	0.04
MFV PH-C3	3	3	0.00	1.00	1	10	−2.98	0.01
PFV PH-C3	0	1	−0.25	0.80	−1	6	−2.33	0.03
EDV PH-C3	−6	4	−1.64	0.11	2	11	−2.24	0.03
PI PH-C3	2	−3	0.82	0.42	−3	−6	0.82	0.42
RI PH-C3	3	−2	1.44	0.16	−1	−2	0.60	0.55

MFV = Mean flow velocity, PFV = Peak flow velocity, EDV = End-diastolic velocity, PI = Pulsatility index, RI = Resistive index

Discussion

Analysis of CBF parameters during short-term exposure to traffic noise revealed no significant changes in both MCA. It was shown in the animal model that moderate acoustic stimulation (broadband noise below 85 dB) induces an increase in metabolic activity in the cochlea and lower auditory centers,[12] while exposure to high-intensity noise (above 100 dB) causes a decrease in metabolic activity in these structures.[13] In human subjects, acoustic stimulation of much lower intensity causes increased resistance in vessels, which was shown for the basilar artery during short exposure to pink noise.[14] To our knowledge, there have been no studies so far investigating the influence of traffic noise on the condition and function of cerebral vessels in the secondary and tertiary acoustic cortexes. These areas are supplied by the MCA, so our study indirectly enables better insight into their reactivity and functioning. Some papers analyzing the relevance of music for CBF have stated that it affects cerebrovascular activity mainly through the autonomic nervous system, including synchronization of HR variability.[151617] The trend for an increase in HR on exposure to noise shown in our study may suggest a similar background.

The performance of PASAT in silence resulted in significantly increased BFVs in both MCA, demonstrating the activation of both hemispheres during this mental task. The more pronounced changes in the left MCA could suggest greater involvement of the dominant hemisphere, which corresponds with data from other reports.[181920] Raised asymmetric CBF (higher in the dominant hemisphere) was observed during speech recognition or processing of acoustic stimuli.[20] However, it has to be stressed that PASAT is a measure of attention and concentration abilities rather than of the understanding or processing of the acoustic stimuli.

We used traffic noise as a distractor during the mental task performance and found that it modified the pattern of flow parameters. The significant changes of flow velocities as well as RI and PI only on the left side again indicated an asymmetric involvement of hemispheres in these conditions. In addition, the strong correlation between MFV and age in the right MCA suggested greater involvement of the right hemisphere in older subjects during excessive load (mental task + distraction). However, it is worth highlighting that the increase in flow velocities on both sides was lower during PASAT performance on exposure to noise than in silence. This may suggest the general restriction of CBF during mental task performance in noise, with preserved left hemisphere dominance but also with greater involvement of the right hemisphere in older participants. Similar evidence for the processing of speech stimuli in different conditions is provided by electrophysiological studies. In individuals exposed to occupational noise for at least 5 years, left hemisphere dominance in speech discrimination becomes right-hemisphere-preponderant.[10] Additionally, the ability to recognize speech in noisy conditions depends on processing acoustic stimuli, as well as on cognitive control — and this is known to deteriorate with age and increasing hearing loss.[2122] However, our findings are not consistent with those from studies based on functional MRI, which revealed the higher activation of left temporal lobe and both cerebellar hemispheres in the young subjects during processing of speech stimuli on exposure to noise in comparison with silent conditions.[23]

The recruitment of additional brain regions in older participants, represented by a more pronounced increase in MFV in the right MCA, could be explained as the process of neuroplasticity aimed at preserving an adequate level of mental performance in more demanding conditions. A similar effect has been observed in neuroimaging and electrophysiological studies and has been described as the “hemispheric asymmetry reduction in older adults” (HAROLD) model, which states that under similar circumstances, cerebral activity during cognitive performance tends to be less lateralized in older than in younger adults. Age-related hemispheric asymmetry reduction may reflect a compensatory function or a dedifferentiation process.[24252627] These relationships certainly deserve further investigation, because the problem of the perception and processing of speech stimuli in noisy conditions is associated not only with neurocognitive but also with social aspects, especially in industrial and urban areas.

The mean results of PASAT in our subjects did not change significantly on exposure to noise (surprisingly, 58% of subjects performed the test even better). The volume of PASAT recording applied concomitantly with noise was increased accordingly so that the noise was supposed to act as a distractor for the attention and not plainly to mask the sounds of the task. However, exposure to noise was used later within the testing session, as were the measures made for the left MCA (also associated with better PASAT results). Repeated performance of PASAT has been shown to improve its results, which may be associated with developing an effective strategy of handling the test as well as diminishing the anxiety and stress caused by the first exposure to this mental task.[28] Overall, this seems to indicate that in our study a learning effect affected predominantly the performance of PASAT in particular conditions. For further investigation into relationships between noise and cognitive functions, it might be useful to apply mental tasks engaging the visual and not the auditory system.

The only finding where male and female subjects differed involved the asymmetry of blood flow changes during the mental task performance. The BFVs in the left MCA were higher in males than in females during performance of PASAT in silence as well as on exposure to noise. In addition, during performance of the mental task in silence, a strong reduction of pulsatility and the RI on the left side was noted only in males, which indicates a higher metabolic demand. All those findings suggest more pronounced engagement of the left hemisphere during PASAT performance in males than in females. These differences may be associated with diverse coping strategies used for the mental task. Such observations have been made for gender differences in patterns of brain activation in functional magnetic resonance imaging (fMRI) studies.[29]

Our study has some limitations, which have to be considered. In order to assess the interhemispheric differences we also analyzed the relative changes of flow parameters during each activation task in comparison to the preceding rest phase, because the flow in both MCA was not measured simultaneously. Repeated exposure to noise during the session might have caused a habituation effect, as had been shown by Bernardi (repeated music samples had a decreasing impact upon subsequently measured CBF in MCA).[15] Nevertheless, we did not notice any such effect on fTCD measures with repeated exposure to noise, in contrast to improving PASAT results with repeated performance. Therefore, we abstained from further analysis of relationships between exposure to noise and cognitive performance.

Conclusion

In conclusion, the results of this study suggest that exposure to noise causes deterioration in CBF during mental task performance. The nondominant hemisphere seems to become more involved in the processing of auditory signals in adverse auditory conditions. These relationships are more pronounced in females and tend to increase with age.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.